THE DESIGN OF ADAPTIVE INTERFACES

FOR ENTERPRISE RESOURCE PLANNING SYSTEMS

Akash Singh and Janet Wesson

Department of Computing Sciences, Nelson Mandela Metropolitan University

P.O. Box 77000, Port Elizabeth, South Africa

Keywords: AUIs, ERP systems, Small enterprises, System design, Usability.

Abstract: Research has shown that AUIs can improve the usability of software systems. Enterprise Resource Planning

(ERP) systems are an example of complex software systems which suffer from several usability issues.

Limited research has been published on the applicability of AUIs to the domain of ERP systems. This paper

aims to address this limitation by discussing the design of AUIs for ERP systems. A secondary contribution

of this paper is the proposal of an adaptive taxonomy for ERP systems. Application of the proposed

taxonomy and the proposed design is demonstrated through the implementation of an AUI for an existing

ERP system, namely SAP Business One (SBO). The AUI was designed to address several of the usability

issues identified in SBO, specifically efficiency and learnability.

1 INTRODUCTION

Adaptive interfaces (AUIs) have the potential to

reduce the complexity of user interfaces (UIs) and

improve the usability of complex systems. This is

achieved by means of adapting the UI automatically

through the use of various types of adaptation and

adaptive components (Letsu-Dake and Ntuen, 2009,

Leung et al., 2006).

One such complex system which suffers from

several usability issues is Enterprise Resource

Planning (ERP) systems (Yeh, 2006). AUIs could

potentially address the usability issues of ERP

systems. Limited evidence exists, however, to

demonstrate how AUIs should be designed for ERP

systems.

This paper aims to address the above limitation

by proposing a design of AUIs for ERP systems.

The remainder of this paper comprises the following

sections: Section 2 contains a discussion on related

work. An adaptation taxonomy for ERP systems is

proposed in Section 3. Section 4 discusses the design

of AUIs for ERP systems. Section 5 presents a

discussion on how the proposed design was

implemented using an existing ERP system. Section

6 contains the conclusions and future work.

2 RELATED WORK

Usability studies of ERP systems indicate that ERP

systems suffer from several usability issues (Topi et

al., 2005). However, most of these studies were

conducted on ERP systems for large enterprises.

Limited research exists on the usability issues of

ERP systems for small enterprises.

A usability study of an ERP system, designed

specifically for small enterprises, was conducted

using SAP Business One (SBO). The usability

evaluation of SBO enabled the identification of 37

usability issues. These usability issues were

classified based on the criteria proposed by Singh

and Wesson (2009).

These usability issues were further analysed to

determine which issues were a result of the UI being

complex. Only 11 of the 37 identified usability

issues related to the complexity of the UI. These 11

usability issues are presented in Table 1.

AUIs could potentially resolve the identified

usability issues through the application of various

types of adaptation. Adaptation can occur in four

basic forms, namely content adaptation, presentation

adaptation, navigation adaptation and device

adaptation (Ramachandran, 2009, Peng and Silver,

2007).

281

Singh A. and Wesson J..

THE DESIGN OF ADAPTIVE INTERFACES FOR ENTERPRISE RESOURCE PLANNING SYSTEMS.

DOI: 10.5220/0003486302810286

In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 281-286

ISBN: 978-989-8425-56-0

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

The research conducted for this paper did not

consider device adaptation as this study focuses on

ERP systems for desktop computers.

Content and presentation adaptation could be

used to address usability issues which relate to task

support and learnability, whilst navigation

adaptation can be applied to usability issues which

relate to navigation. The mapping of the various

adaptation types to the identified usability issues is

presented in Table 1.

3 PROPOSED ADAPTATION

TAXONOMY

An adaptation taxonomy was proposed by Knutov et

al. (2009) to illustrate the various methods,

techniques and approaches that could be used to

implement the various types of adaptations for

hypermedia systems. A detailed literature study

revealed that such a taxonomy does not exist for

ERP systems. This section proposes an adaptation

taxonomy for ERP systems that can be used to guide

and support the design and implementation of AUIs

for ERP systems.

The proposed taxonomy (Figure 1) presents

several modifications over the original taxonomy

proposed by Knutov et al. (2009). Components of

the original taxonomy are illustrated in blue and the

new proposed components are illustrated in orange.

Only those methods that could be used to address the

usability issues identified in Table 1 were selected

from the original taxonomy.

The approaches and techniques originally

proposed for hypermedia systems were replaced by

existing HCI design patterns for desktop

applications. Existing HCI patterns were selected to

support the design of usable ERP systems. These

patterns can assist in solving common UI design

problems within a given design domain (Tidwell,

2006).

Several existing catalogues of HCI patterns exist

for both desktop and web-based systems. The aim of

these catalogues is to improve system usability by

providing patterns to support UI and interaction

design.

The most comprehensive, recent and widely

referenced catalogues include Tidwell (2006) and

van Welie (2008). The HCI patterns proposed by

Tidwell (2006) were selected due to their

comprehensiveness and applicability to desktop

systems.

The following sections discuss how the proposed

HCI patterns could be used to implement the various

types of adaptation.

Figure 1: Proposed ERP Adaptation Taxonomy.

3.1 Content Adaptation

Based on the proposed ERP adaptation taxonomy,

content adaptation can be implemented by inserting,

removing, or altering fragments of information. The

aim of inserting, removing or altering fragments of

information is to ensure efficiency and to improve

user productivity.

3.2 Presentation Adaptation

Several presentation adaptation techniques are

proposed (Figure 1) to address the usability issues

that were identified with regards to task support,

learnability and customisation.

Removing fragments uses the techniques of

responsive disclosure and extras on demand. These

two techniques aim to reduce the number of controls

on the UI by only displaying what is necessary at a

particular point in time whilst providing an option to

display more or everything (extras-on-demand).

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

282

Table 1: Usability Issues of SBO and Relevant Adaptation Types.

No. Usability Issue Category Adaptation Type

1. Too many steps required to complete a task Task Support Presentation

2. SBO does not enable efficient completion of tasks Task Support Content

3. SBO does not improve user productivity Task Support

Content and

Presentation

4. SBO does not automate routine tasks Task Support Content

5. Finding information and functionality is difficult Navigation Navigation

6. SBO cannot guide the user through the correct sequence of tasks Navigation Navigation

7. Evidence of the next sequence of steps to complete a task is not provided Navigation Navigation

8. There are too many unused fields on the screen Presentation Presentation

9. Layout of the UI does not contribute to the efficient completion of tasks Presentation Presentation

10.

Steps need to be manually recorded the first time in order to be

remembered for future use

Learnability Presentation

11. Personalisation of UIs are not possible Customisation

Content and

Presentation

The method of disabling fragments uses the

technique of responsive enabling. This technique

aims to address the issue of guiding the user through

a task by enabling task activities in an ordered

manner.

Sorting fragments uses two techniques, namely

good defaults and movable panels. This method

addresses the usability issue of UI personalisation.

This method will re-order the fragments of the UI

based on the user’s previous interactions (good

defaults) and could also allow the user to re-order

the fragments based on his/her own preference

(movable panels).

3.3 Navigation Adaptation

Most of the usability issues relating to navigation

identified in Table 1 relate to guidance. For this

reason the guidance adaptation method was selected

from the original taxonomy.

The following techniques are proposed which

aim to ensure that users are guided through the

correct sequence of tasks (sequence map); evidence

of the next sequence of steps are provided (sequence

map and colour-coded sections); and finding

information and functionality is facilitated (multiple-

level help).

4 DESIGN

This section proposes a design for AUIs for ERP

systems. The proposed design is made up of two key

components: system architecture and model design.

4.1 System Architecture

The proposed architecture in Figure 2 is an

extension of an existing AUI system architecture

proposed by Ramachandran (2009). Components of

the original architecture are illustrated in blue in and

the new components are illustrated in orange (Figure

2). The architecture proposed by Ramachandran

(2009) was selected as it was general enough to be

specialised to the domain of ERP systems, it

supported presentation and navigation adaptation; it

included a user model; and it separated the business

logic of the system from the actual UI.

The proposed system architecture (Figure 2)

differs from the original architecture through the

addition of the following components:

 An inference engine (monitoring the user

interactions & updating the AUI);

 Support for content adaptation (missing from

the original architecture);

 An ERP engine (contains both the business

logic and the ERP system logic); and

 Support for the ERP system database tables

and application data tables within the

database.

These additions are implemented across three

different sections within the proposed system

architecture: the database layer, the application layer

and the presentation layer.

4.1.1 Database Layer

The database layer contains the central database.

This database is accessed by the ERP system and is

THE DESIGN OF ADAPTIVE INTERFACES FOR ENTERPRISE RESOURCE PLANNING SYSTEMS

283

responsible for storing the default ERP system

tables, any customised tables’ specific to a particular

enterprise, application data, adaptation rules and the

AUI components (user model, task model and dialog

model).

4.1.2 Application Layer

The application layer is responsible for the

functionality of the ERP system and is also

responsible for the adaptation and generation of the

UI. The application layer makes use of an inference

engine, an adaptation engine and an ERP engine.

The inference engine is responsible for

monitoring any interaction that the user has with the

AUI and for generating and updating the UI. The

inference engine was not part of the original system

architecture.

A log file is stored for each user, in order for the

inference engine to successfully monitor and capture

the user’s interaction with the UI.

The adaptation engine takes the input from the

inference engine and determines what type of

adaptation to perform on the UI in order to improve

user productivity. The proposed system architecture

is capable of supporting content, presentation and

navigation adaptation, whilst the original

architecture only supported presentation and

navigation adaption.

The ERP engine is the existing functional ERP

system that controls all of the business logic and

ERP functionality. The adaptive engine will interact

with the ERP engine in order to maintain the

integrity of the system and the data. The ERP engine

was not part of the original architecture

(Ramachandran, 2009) and was included in order to

specialise the proposed architecture for ERP

systems. Inclusion of the ERP engine also indicates

how the adaptive engine interacts with the ERP

system.

4.1.3 Presentation Layer

The presentation layer contains the AUI. The AUI

will be generated based on one or more of the

required adaptation types from the adaptation engine

and delivered through the inference engine.

Any interaction performed by the user will be

monitored through the inference engine.

Determining the validity of the data will be done by

the ERP engine and if accepted, this data will be

saved to the database.

4.2 Adaptive Components

The various types of adaptation discussed in Section

2 make use of several adaptive components in order

to execute the adaptation. These adaptive

components are represented in the form of

declarative models.

4.2.1 User Model

User models are a critical component of any AUI as

they provide the information necessary to adapt the

user model between the different users (Tran et al.,

2009).

The proposed user model will contain a summary

of the key elements that the ERP system will require

in order to adapt the UI. These elements will be

grouped according to the various tasks performed by

the user.

The data stored within the user model will aid

the decisions made by the adaptation engine in terms

of supporting presentation adaptation and content

adaptation. Data stored within the user model

includes the sequence in which controls were

selected for a particular task. This could assist with

presentation adaptation in terms of re-ordering the

controls on the UI. All of the data stored within the

user model will be continuously updated based on

the users’ interaction with the ERP system.

4.2.2 Task Model

Task models can be defined as the hierarchical

representation of the tasks and sub-tasks (activities)

performed by users of a particular software system.

The proposed task model is based on a generic

description of task models and contains:

 The name, goal and form ID of the task;

 The pre- and post-conditions of the task; and

 The related sub-tasks (activities).

The task model can assist with adaptation by

means of supporting navigation adaptation in terms

of:

 Guiding users through the correct sequence of

steps (activities); and

 Providing evidence of the next sequence of

steps declared within the task model.

Manipulation of tasks is done by means of a

dialog model and a presentation model, which are

discussed in the following sections.

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

284

Figure 2: Proposed System Architecture.

4.2.3 Dialog Model

A Dialog model complements the task model by

means of describing and organising various task

activities that need to be performed on the UI. The

proposed dialog model is based on the description of

the generic dialog model and will support the

process of presentation adaptation (Figure 1) by

means of visually:

 Sorting fragments and the controls within

these fragments; and

 Enabling and disabling fragments.

The values of the various controls per activity

could change depending on the user’s interaction

with the UI.

5 PROTOTYPE

The proposed design and adaptation taxonomy was

implemented using SBO. SBO was selected as it is

widely used by small enterprises, it was the most

extensible ERP system, and it allowed for the

separation of business logic and AUI logic.

Further analysis of the usability issues presented

in Table 1 revealed that these issues could be

classified as either efficiency issues (to improve task

support through better task completion times) or

learnability issues (how quickly and effectively can

a novice SBO user learn how to use the ERP

system). This section will briefly discuss how the

various types of adaptation were implemented in

order to improve the efficiency and learnability of

SBO.

Content adaptation was implemented using list-

based adaptation techniques. This was applied the

dropdown boxes of SBO using the frequency and

recency algorithms proposed by Gajos et al. (2006).

Presentation adaptation techniques (Figure 1) were

applied to the fragments (a collection of controls

within a grouped section of a form) in SBO.

Fragments were either disabled or enabled (disabling

and enabling fragments pattern) based on the

sequence order in the task and dialog models.

Navigation adaptation was implemented using the

sequence map pattern (Figure 1) in order to provide

improved guidance. This was done by introducing

an additional colour state (green – to symbolise

appropriate data entry) to the controls, in addition to

the existing colour states of the controls (white –

enabled; grey – disabled; and orange – active) to

support guidance and task completion.

6 CONCLUSIONS AND FUTURE

WORK

ERP systems currently suffer from several usability

issues which are caused by complex UIs. Research

has shown that AUIs can improve the usability of

software systems, but limited guidance exists to

support the design of AUIs for ERP systems. This

paper has proposed an adaptation taxonomy for ERP

systems and the design of an AUI for ERP systems

THE DESIGN OF ADAPTIVE INTERFACES FOR ENTERPRISE RESOURCE PLANNING SYSTEMS

285

using this taxonomy.

A prototype AUI was successfully implemented

using SBO. The development of this prototype

provided evidence to support the potential benefits

of AUIs to improve the usability of ERP systems

Future work will include conducting a comparative

user study to determine if the usability of the AUI is

better than the non-adaptive version.

ACKNOWLEDGEMENTS

The support of SAP Research Pretoria and SAP

Meraka UTD (CSIR) towards this research is hereby

acknowledged. Opinions expressed and conclusions

arrived at are those of the authors and not

necessarily to be attributed to the companies

mentioned in this acknowledgement.

REFERENCES

Gajos, K. Z., Czerwinski, M., Tan, D. S. & Weld, D. S.

(2006). Exploring the design space for adaptive

graphical user interfaces. Proceedings of the working

conference on Advanced visual interfaces. Venezia,

Italy: ACM.

Knutov, E., De Bra, P. & Pechenizkiy, M. (2009). AH 12

years later: a comprehensive survey of adaptive

hypermedia methods and techniques. New Review of

Hypermedia and Multimedia, 15, 5-38.

Letsu-Dake, E. & Ntuen, C. A. (2009). A Conceptual

Model for Designing Adaptive Human-Computer

Interfaces Using the Living Systems Theory. Systems

Research and Behavioral Science, 26, 15-27.

Leung, A., Morisson, S., Wringe, M. & Zou, Y. (2006).

Developing an adaptive user interface in eclipse. In

Proc. the Eclipse Technology Exchange Workshop at

European Conference on Object Oriented

Programming. Nantes, France.

Peng, X. & Silver, D. L. (2007). Interface Adaptation

Based on User Expectation. Proceedings of the 21st

International Conference on Advanced Information

Networking and Applications Workshops - Volume 02.

IEEE Computer Society.

Ramachandran, K. (2009). Adaptive User Interfaces for

Health Care.

Singh, A. & Wesson, J. L. (2009). Evaluation criteria for

assessing the usability of ERP systems. In: In

Proceedings of the 2009 Annual Research Conference

of the South African institute of Computer Scientists

and information Technologists (SAICSIT '09),

October 12 - 14 2009 Vanderbijlpark, Emfuleni, South

Africa. ACM, 87-95.

Tidwell, J. (2006). Designing Interfaces, O'Reilly Media,

Inc.

Topi, H., Lucas, W. & Babaian, T. (2005). Identifying

Usability Issues with an ERP Implementation. In:

International Conference on Enterprise Information

Systems, 24-28 May 2005 Miami, USA.

Tran, V., Vanderdonckt, J., Kolp, M. & Faulkner, S.

(2009). Generating User Interface from Task, User and

Domain Models. The Second International Conference

on Advances in Human-Oriented and Personalised

MEchanisms, Technologies and Services (CENTRIC

'09). Porto, Portugal.

Van Welie, M. (2008). A Pattern Library for Interaction

Design. Available: http://www.welie.com/ [Accessed

July 15 2010].

Yeh, J. Y. (2006). Evaluating ERP Performance from User

Perspective. In: IEEE Asia-Pacific Conference on

Services Computing (APSCC'06), December 12-15

2006. IEEE Computer Society, 311-314.

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

286